A novel data-based quality-related fault diagnosis scheme for fault detection and root cause diagnosis with application to hot strip mill process

In this paper, a new technology or solution of quality-related fault diagnosis is provided for hot strip mill process (HSMP). Different from traditional data-based fault diagnosis methods, the alternative approach is focused more on root cause diagnosis. The new scheme addresses the quality-related fault detection with the developed modified canonical variable analysis (MCVA) model, then the advantage of original generalized reconstruction based contribution (GRBC) is followed to identify the faulty variables. Meanwhile, a new transfer entropy (TE)-based causality analysis method is proposed for root cause diagnosis of quality-related faults. Finally, the whole proposed framework is practiced with real HSMP data, and the results demonstrate the usage and effectiveness of these approaches.